Article pubs.acs.org/jnp
Metabolites of Microcystis aeruginosa Bloom Material from Lake Kinneret, Israel Marina Lifshits and Shmuel Carmeli* Raymond and Beverly Sackler School of Chemistry and Faculty of Exact Sciences, Tel-Aviv University, Ramat Aviv, Tel-Aviv 69978, Israel S Supporting Information *
ABSTRACT: Six new metabolites, micropeptin KT1042, microguanidine KT636, aeruginosins KT608A, KT608B, and KT650, and pseudoaeruginosin KT554, were isolated along with the known micropeptins SF909 and HM978, cyanopeptolin S, anabaenopeptin F, and the two isomers of planktocyclin-S-oxide from a bloom material collected from Lake Kinneret, Israel, in March 2005. The structure elucidation and biological activity of the six new natural products isolated from this bloom material and the related aeruginosin GH553 are described.
■
M
RESULTS AND DISCUSSION The bloom material was separated in the laboratory to an aqueous solution and cells by aggregation with alum. The cells were immediately frozen after separation and freeze-dried. The aqueous solution (12 L) was passed through an Amberlite XAD-2 column and extracted thereafter with MeOH. The MeOH extract, which inhibited trypsin and chymotrypsin, was flash chromatographed on a reversed-phase C18 column. The fractions that exhibited protease inhibitory activity were further separated by gel filtration on a Sephadex LH-20 column and reversed-phase HPLC columns to afford six pure natural products: micropeptins KT1042 (1), HM978,13 and SF90914 and aeruginosins KT608A (3), KT608B (4), and KT650 (5). The freeze-dried cyanobacterium biomass was extracted with 70% MeOH in H2O. The cell extract was separated in the same way as the aqueous extract to afford micropeptins KT1042 (1), HM978,13 and SF909,14 cyanopeptolin S,15 microguanidine KT636 (2), aeruginosins KT608A (3), KT608B (4), and KT650 (5), pseudoaeruginosin KT554 (6), anabaenopeptin F,16 and the two S-oxide isomers of planktocyclinS-oxide (see Figure S50 in the Supporting Information).17,18 Micropeptin KT1042 (1) was isolated as a colorless, glassy material. It presented a sodiated MALDI TOF quasimolecular ion at m/z 1065.4971, in agreement with the molecular formula C53H70N8NaO14. The NMR spectra of 1 in DMSO-d6 (Table 1) revealed its peptidic nature and that it belongs to the micropeptins, i.e., nine carbonyl carbons in the 13C NMR spectrum, five secondary amide doublet and two primary amide singlet protons, two hydroxy protons resonating at δH 6.04 (brd) and 5.51 (brs), and an NMe group δH 2.56 (s), in the 1H NMR spectrum. Taking into account the characteristic NMe-aromatic amino acid and the N,N-disubstituted-amino acid of the micropeptins, the five amide doublet protons revealed that this micropeptin was composed of seven amino acids. The existence
arine and freshwater bloom-forming genera of cyanobacteria are well known for the structural diversity of their secondary metabolites and for the high structural variation within each particular group of metabolites.1,2 Freshwater toxic cyanobacteria are among the most studied genera of cyanobacteria due to their serious threat to human health.3 More than 80 different members of the hepatotoxic microcystins, which are produced by water-bloom-forming genera of cyanobacteria, are known to date.4 They are usually accompanied by the micropeptins, a group of serine protease inhibitors that includes more than 130 members,5 the anabaenopeptins (38 isolated variants),6 the aeruginosins (19 isolated variants),7 the microginins (34 isolated variants),8 the microviridins (15 isolated variants),9 and many other compounds that are not grouped together. All of these compounds are peptidic in nature, but they usually contain certain structural elements that interfere with their hydrolysis by proteolytic enzymes, such as β-amino acids (aa’s), long chain aa’s, dehydro aa’s, modified aa’s, iso-linked aa’s, N-methylated aa’s, D-aa’s, and other isomeric aa’s. In certain cases allo-amino acids are incorporated into the backbone of these peptides. Most of the aeruginosins belong to one group that contains the proline-mimicking amino acid L-2-carboxy-6-hydroxyoctahydroindole (L-Choi),7 while only one, aeruginosin EI461, contains the L-diepi-Choi.10,11 As part of our ongoing research on the chemistry and chemical ecology of cyanobacterial blooms in water bodies, the biomass of bloom material (TAU strain IL-347) of Microcystis aeruginosa collected from Lake Kinneret, Israel, in March 2005 was chemically investigated. The extract of this bloom afforded some unusual isomeric compounds belonging to the micropeptins and aeruginosins. All of the isolated aeruginosins contain a new (2R,3aR,6R,7aR)Choi (D-3a,7a-diepi-Choi) moiety. The isolation and structure elucidation of the new secondary metabolites isolated from this cyanobacterial bloom biomass and the related aeruginosin GH553,12 their biological activities, and the biogenesis of this new group of aeruginosins are discussed below. © 2012 American Chemical Society and American Society of Pharmacognosy
Received: November 16, 2011 Published: January 26, 2012 209
dx.doi.org/10.1021/np200909x | J. Nat. Prod. 2012, 75, 209−219
Journal of Natural Products
Article
Table 1. NMR Data of Micropeptin KT1042 (1) in DMSO-d6a position Val 1 2
δC, mult.b 172.4, C 56.4, CH
δH, mult., J (Hz)
NOE correlationsd
3 4
30.9, CH 19.4, CH3
2.05, m 0.83, d (6.7)
Val-1,3,4,5, NMePhe1 Val-4,5 Val-2,3
5
17.8, CH3
0.72, d (6.7)
Val-2,3
7.60, d (9.5)
NMePhe-1
5.05, dd (11.5,2.0) 2.60, m
NMePhe-3, NMe NMePhe2,4,5,5′ NMePhe-2
NMePhe-5,5′,6,6′, Val-NH, Ile-2 NMePhe3b,5,5′,6,6′ NMePhe-3a,5,5′,6,6′
NMePhe3,7 NMePhe-4
NMePhe-2,3, Ile2,4,5 NMePhe-2,3, Ile2,4,5
NH
NMePhe 1 2 3
169.3, C 60.7, CH 34.5, CH2
4.70, dd (9.6, 5.2)
HMBCcorrelationsc
3.35, m 4 5, 5′
137.7, C 129.7, CH
7.20, m
6, 6′
128.9, CH
7.25, m
7
126.6, CH
7.15, m
NMe Ile 1 2
30.5, CH3 170.1, C 55.2, CH
2.56, s
4.37, d (10.4)
3
34.4, CH
1.64, m
4
24.5, CH2
0.13, ddq (13.5,6.8,7.2)
5 6 Ahp 2 3 4 5
12.2, CH3 14.1, CH3 169.5, C 49.2, CH 21.8, CH2 29.8, CH2
−0.26, ddq (13.5,9.6,7.2) 0.53, t (7.2)
NMePhe5,5′ NMePhe-2, Ile-1 Ile-1,3,4,6, Ahp-6 Ile-5,6 Ile-5,6
Ile-2,3,4
0.45, d (6.7)
Ile-3,4
4.46, 1.72, 2.51, 1.71, 1.83,
Ahp-2,4
m m m m m
position
Val-3,4,5,NH
Val-2,4,5,NH Val-2,3,NH, NMePhe-NMe Val-2,3,NH, Ahp-6OH, NMePheNMe Val-2,3,4,5, Ahp-6OH, NMePhe-2, NMe
6 6-OH NH
74.2, CH
Tyr 1 2 3
170.2, C 54.0, CH 35.5, CH2
4 5, 5′ 6, 6′ 7 7-OH NH
128.4, C 129.8, CH 115.2, CH 156.7, C
Thr 1 2
169.0, C 54.7, CH
3 4 NH Gln 1 2 3
Val-4,5,NH
4 5 NH
Ile-3,4,6, NMePhe2,5,5′,6,6′ Ile-2,4a,5,6, Ahp-6, NPhe-5,5′,6,6′ Ile-2,3,4b, NMePhe-5,5′,6,6′, Ile2,4a,5,6, NMePhe-5,5′,6,6′
δC, mult.b
Ile-3,4, NMePhe5,5′,6,6′ Ile-2,3,4
2-OH 3
Ahp-4a,NH Ahp-3 Ahp-6-OH,NH Ahp-6 Ahp-6
4 5, 5′ 6, 6′ 7 7-OH
4.90, brs 6.04, brs 7.35, d (9.0)
72.1, CH 17.7, CH3 171.8, C 51.6, CH 28.6, CH2 31.5, CH2 174.3, C
40.0, CH2
128.6, C 130.4, CH 114.9, CH 156.5, C
Ahp-4 Ahp-3, Tyr1
NOE correlationsd Ahp-5,6-OH, Ile-3 Ahp-4b,6, Val-5,NH Ahp-3,4b, Thr-3, Tyr-NH
Tyr-1,3 Tyr-2,4,5,5′
Tyr-5,5′,NH Tyr-3b,5,5′,NH Tyr-3a,5,5′,NH
6.94, d (8.0) 6.58, d (8.0)
Tyr-3,7 Tyr-4,7
Tyr-2,3,6,6′ Tyr-5,5′,7-OH
Tyr-2, Thr1
4.57, d (9.4)
Thr-1, Val-1
5.48, q (6.5) 1.15, d (6.5) 7.98, d (9.4)
Thr-4, Val-1 Thr-2,3 Gln-1
4.45, m
Gln-1,3,4
1.73, m 1.83, m 2.10, m
Gln-1,2,4,5
7.78, d (8.0)
Gln-2, Hpla1 Gln-4
6.79, brs, 7.14, m 173.5, C 72.7, CH
HMBCcorrelationsc
4.45, m 3.30, m 2.50, m
9.13, brs 8.51, d (8.7)
5-NH2 Hpla 1 2
δH, mult., J (Hz)
Gln-2,3,5
4.03, brd (7.5) 5.51, brs 2.65, dd (14.0, 3.2) 2.49, m
Hpla-2 Hpla1,2,5,5′
7.00, d (8.0) 6.62, d (8.0)
Hpla-3,7 Hpla-4,7
Tyr-6,6′ Thr-2,3, Ahp-NH
Thr-3,4,NH, TyrNH Tyr-2,NH, Ahp-NH Thr-2,NH Thr-2,4, Gln-2 Gln-3,4,1-NH, ThrNH Gln-2,1-NH Gln-2,1-NH Gln-2,1-NH Gln-2,3,4,Hpla-2,2OH
Hpla-3,5,5′,2-OH, Gln-1-NH Hpla-2, Gln-1-NH Hpla-2,3b,5,5′ Hpla-2,3a,5,5′ Hpla-2,3,6,6′ Hpla-5,5′
9.13, brs
Table 1. continued
a 500 MHz for 1H, 125 MHz for 13C. bMultiplicity and assignment from an HSQC experiment. cHMBC correlations, optimized for 8 Hz, are from the proton(s) stated to the indicated carbon. dSelected NOEs from ROESY experiment.
correlations with protons of an N-methyl resonating at δH 2.56 and the protons of methylene-3. HMBC correlations of C-3 and H-3 with the phenyl protons and carbons, respectively, established the connectivity of the phenyl ring to the aliphatic chain, and correlation of H-2 with the carbonyl, which resonated at δC 169.3, completed the structure elucidation of the NMePhe moiety. The doublet proton that resonated at δH 4.37 was assigned as the α-proton of an isoleucine moiety on the basis of the COSY and TOCSY correlations, but failed to present any correlation with either an amine or amide proton. The carbonyl of the Ile δC 170.1 was assigned on the basis of its HMBC correlation with Ile-H-2. The correlation of the latter proton with the aminal carbon (AhpC-6) supported the establishment of this amino acid moiety as N,N-disubstituted-Ile. The sequence of these residues in 1, ValNMePhe-N,N-disubstituted-Ile-Ahp-Tyr-Thr-Gln-Hpla, and the
of the amino hydroxy piperidone (Ahp) and the p-hydroxyphenyllactic acid (Hpla) was suggested on the basis of the COSY and HSQC spectra, which presented the hydroxy proton at δH 6.04 coupled to the aminal proton at δH 4.90 (brs) (δC 74.2, CH) of the Ahp moiety, and the hydroxy proton δH 5.51 coupled to the oxymethine at δH 4.03 (brd) (δC 72.7, CH) for Hpla. Interpretation of the COSY, TOCSY, HSQC, and HMBC data (Table 1) allowed the assignment of the structure of eight residues: Val, NMePhe, N,N-disubstituted-Ile, Ahp, Tyr, Thr, Gln, and Hpla. The structure determination of the NMePhe moiety started with H-2, which resonated at δH 5.05 (dd) and was coupled through COSY correlations to a pair of benzylic protons (H-3,3′, δH 2.60 and 3.35). An HSQC correlation established the connectivity of H-2 with an amino-methine carbon resonating at δC 60.7. The latter carbon exhibited HMBC 210
dx.doi.org/10.1021/np200909x | J. Nat. Prod. 2012, 75, 209−219
Journal of Natural Products
Article
substituted-Ile in both compounds is significantly different, suggesting a different conformation or configuration at C-3. Because the L-FDAA reagent does not allow the discrimination of Ile from allo-Ile on a reversed-phase HPLC column, we repeated the analysis with the leucine amide derivative L-FDLA, which established the latter amino acid as L-allo-Ile. This established the structure of micropeptin KT1042 as 1. Comparison of the proton and carbon chemical shifts of all N,Ndisubstituted-Ile moieties in known micropeptins revealed that micropeptin KT1042 (1) is, so far, the only micropeptin that incorporates allo-Ile at this position. This finding also allowed concluding that, in the micropeptins, if an Ile is situated between the Ahp and the NMe-aromatic acid, the proton signal of 6-CH3 will be shifted upfield of TMS, while in the case of allo-Ile, one of the 4-CH2 protons will be shifted upfield of TMS. Because the conformation of the micropeptins in solution is quite rigid (due to the intramolecular hydrogen bonds between 1Val-CO and 4AhpNH and 1Val-NH and 4Ahp-6-OH), there is a significant difference in the carbon chemical shifts, as was observed by us previously, for these two amino acids when they occupy the C-terminus position of the micropeptins.21 Microguanidine KT636 (2) was isolated as a colorless oil. It presented a negative HRESIMS pseudomolecular ion at m/z 635.1002, which matched the molecular formula, C19H32N4O14S3. The 1H NMR spectrum of 2 in DMSO-d6 (Table 2) was relatively broad, displaying signals characteristic of a 1,2,4-trisubstituted-phenyl ring (δH 7.16 dd, 7.24 d, 7.38 brd), five protons adjacent to electronegative substituents [δH 5.18 brs (2H), 4.93 d, 4.85 d, 4.14 d], a nine-proton singlet (δH 3.04), and a broad doublet methyl group (δH 1.18), reminiscent
cyclization of the valine carbonyl to the threonine oxymethine could be independently assigned by the inter-residual HMBC or ROESY correlations (see Table 1). The relative configuration of the Ahp residue, 3S*,6R*, was assigned on the basis of the similarity of its proton and carbon chemical shifts to those of micropeptin HM978 and the NOE of the pseudoaxial-H-4 (H-4pax, δH 2.51) with the 6-OH.13 Marfey’s analysis19 (using L-FDAA as derivatizing reagent) preceded and not preceded by Jones’ oxidation,20 established the L-configuration of Glu × 2 (from Gln and from 3S,6R-Ahp), Ile, NMePhe, Thr, Tyr, and Val, while analysis on a chiral-phase HPLC-column established the absolute configuration of Hpla as L. Comparison of the proton chemical shifts of the N,N-disubstituted-Ile of 1 with those of micropeptin HM978, both sharing similar chemical environments (Figure 1), revealed
Figure 1. Comparison of the proton and carbon chemical shifts of the isoleucine moieties of micropeptins HM978 and KT1042 (1).
that the anisotropic effect of the neighboring aromatic ring is observed on the opposite arms of the isoleucine side chains in both compounds. The carbon chemical shifts of the N,N-di211
dx.doi.org/10.1021/np200909x | J. Nat. Prod. 2012, 75, 209−219
Journal of Natural Products
Article
for 4. Full assignment of the 1H and 13C NMR data suggested that they presented opposite configurations at the stereogenic center of their Hpla moiety. Aeruginosin KT608A (3) exhibited a positive HRESIMS molecular adduct ion at m/z 609.3409 [M + H]+, consistent with the molecular formula C32H45N6O6 and 14 degrees of unsaturation. The interpretation of the NMR data in DMSO-d6 was rather complicated due to the high similarity of the chemical shifts of the proton and carbon signals of both rotamers. The 1H NMR spectrum of 3 presented three pairs of amide protons (a pair of doublets and two pairs of triplets), two pairs of signals indicating the presence of a para-substituted phenol and a phenyl in the aromatic region, and four pairs of methine protons next to electronegative atoms in the 4.9−3.9 ppm region (Figure S3). The aliphatic region was too complicated to be interpreted. The 13C NMR spectrum revealed three pairs of amide/ester carbonyls, two pairs of guanidine/ phenol sp2 carbons, 14 signals of aromatic sp2 carbons consistent with two pairs of phenyl and phenol moieties, five pairs of methines between 73 and 50 ppm (Figure S4), and several additional methine and methylene signals in the aliphatic region. Both rotamers were fully characterized (Table 3, for the NMR data of the cis rotamer, and Table S3a in the Supporting Information for the trans rotamer), but for clarity only the structure elucidation of the major cis rotamer is discussed below. Interpretation of the data from the COSY, TOCSY, HSQC, and HMBC 2D NMR experiments allowed the assignment of the agmatine, Phe, and Hpla moieties (Table 3). The structure elucidation of the 2-carboxy-6-hydroxyoctahydroindole (Choi) moiety was more challenging. COSY and TOCSY correlations allowed the construction of the sequence of H-2 through H-7a, and the correlations in the HSQC spectrum allowed the assignment of the carbons bearing these protons (Table 3). H-2 exhibited COSY correlations with the adjacent protons resonating at δH 2.38 (a 9.5 Hz coupling constant) and 1.70 (
